Page 103 - AIH-1-4

Artificial Intelligence in Health

ORIGINAL RESEARCH ARTICLE
An exploratory study on the potential of

ChatGPT as an AI-assisted diagnostic tool for
visceral leishmaniasis

Paulo Adriano Schwingel 1,2,3,4† * , Dino Schwingel 1,2† ,
Samuel Ricarte de Aquino 1,5† , Aline Rafaela Soares da Silva 1,2,3 ,
Pedro Paulo Ramos da Silva 1,2 , Renato Augusto da Cruz Pereira 1,2,6 ,
Daniela Conceição Gomes Gonçalves e Silva 1,2,4 ,
Amanda Alves Marcelino da Silva 1,2,3,4 ,
Flavia Emília Cavalcante Valença Fernandes 1,2 ,
Maria Jacqueline Silva Ribeiro 1,2,6 , Paulo Ditarso Maciel Júnior 1,7 ,
Paulo Gustavo Serafim de Carvalho 1,8 , Ricardo Kenji Shiosaki 1,2 ,
Rogério Fabiano Gonçalves 1,2 , Bruno Bavaresco Gambassi 1,2,6 ,
and Paula Andreatta Maduro 1,2,4
1 AI-assisted Diagnostics Research Group, Universidade de Pernambuco, Petrolina, Pernambuco,
Brazil
2 Human Performance Research Laboratory, Universidade de Pernambuco, Petrolina, Pernambuco,
Brazil

Abstract
† These authors contributed equally
to this work.
Visceral leishmaniasis (VL) is a severe parasitic disease that poses significant diagnostic
*Corresponding author: challenges due to its complex presentation and the necessity for comprehensive
Paulo Adriano Schwingel
(paulo.schwingel@upe.br) diagnostic methods. This exploratory study investigates the potential of Chat
Generative Pre-trained Transformer (ChatGPT)/GPT-4, an artificial intelligence (AI)
Citation: Schwingel PA,
Schwingel D, de Aquino SR, chatbot, in assisting the diagnostic process for VL. We evaluated the diagnostic
et al. An exploratory study on accuracy of ChatGPT/GPT-4 in generating differential diagnosis lists for eight clinical
the potential of ChatGPT as an vignette cases of VL, authored by a Brazilian infectious disease doctor. Our findings
AI-assisted diagnostic tool for reveal that ChatGPT/GPT-4 included VL in the top five differential diagnoses in 75%
visceral leishmaniasis. Artif Intell
Health. 2024;1(4):97-106. of the cases (95% confidence interval [CI]: 40.1 – 93.7%) and identified VL as the top
doi: 10.36922/aih.3930 diagnosis in 50% of the cases (95% CI: 30.3 – 86.5%). These results underscore the
Received: June 13, 2024 high potential of ChatGPT/GPT-4 as an AI-assisted diagnostic tool, which is capable
of providing accurate differential diagnoses and assisting healthcare professionals
Accepted: September 20, 2024
in resource-limited settings. The study highlights the broader applicability of AI
Published Online: October 16, 2024 chatbots in medical diagnostics, not only for common conditions but also for
Copyright: © 2024 Author(s). specialized and less prevalent diseases like VL. By integrating AI tools into the
This is an Open-Access article diagnostic workflow, healthcare providers can enhance their diagnostic accuracy and
distributed under the terms of the
Creative Commons Attribution efficiency, ultimately improving patient outcomes. This research contributes to the
License, permitting distribution, growing body of evidence supporting the utility of AI in healthcare and underscores
and reproduction in any medium, the need for further studies to validate these findings across larger and more diverse
provided the original work is
properly cited. clinical scenarios.
Publisher’s Note: AccScience
Publishing remains neutral with Keywords: Tropical neglected diseases; Artificial neural network; Differential diagnosis;
regard to jurisdictional claims in
published maps and institutional Artificial intelligence-assisted diagnosis; Healthcare technology
affiliations.

Volume 1 Issue 4 (2024) 97 doi: 10.36922/aih.3930